Across 18 comparable benchmarks, Kimi K2.6 has better reported results on 15 and worse on 3.
Moderate evidence18 of 18 shared results countedHow we compare
Price basis: International rates vs standard rates. The price section below shows the applicable periods.
| Fact | Kimi K2.6 | GPT-5 |
|---|---|---|
| Context window | 256K tokens | 400K tokens |
| Input / 1M | $0.950 | $1.25 |
| Output / 1M | $4.00 | $10.00 |
| Blended (4:1) | $0.991 | $2.19 |
| Status | Available | Deprecated |
| Released | April 20, 2026 | August 7, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 13 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
Kimi K2.6 ahead on 4 of 4 counted
Kimi K2.6 ahead on 2 of 2 counted
Split 2–2 of 4 counted
Kimi K2.6 ahead on 1 of 1 counted
GPT-5 ahead on 1 of 1 counted
Kimi K2.6 ahead on 1 of 1 counted
Kimi K2.6 ahead on 3 of 3 counted
No shared comparable results for Abstract reasoning, Computer & tool use, Cybersecurity.
Published scores, with matching modes where available. Gaps show Kimi K2.6 relative to GPT-5. Reading the results
| Benchmark | Kimi K2.6 | GPT-5 | Gap |
|---|---|---|---|
| SWE-Bench Pro - PublicRepository Engineering · Accuracy | 58.6Thinking · tools | 41.8≠Thinking High · tools | +16.8 pts#22 vs #59 of 66 |
| SWE-bench VerifiedRepository Engineering · Accuracy | 80.2Thinking · tools | 72.8≠Thinking High · no tools | +7.4 pts#15 vs #49 of 108 |
| Terminal Bench HardAgentic Development · Accuracy | 43.9Thinking · tools | 37.9≠Thinking Medium · tools | +6.0 pts#23 vs #38 of 157 |
| Vibe Code Bench v1.1Code Generation & Editing · Pass rate (%) | 37.9Unspecified Reasoning · tools | 20.1≠Thinking High · tools | +17.8 pts#74 vs #88 of 121 |
| FrontierMath Tier 4 v2Mathematics · Accuracy (verification_code) | 25.6Reported best | 22.0≠Thinking High · tools | +3.7 pts#46 vs #49 of 73 |
| FrontierMath v2Mathematics · Accuracy (verification_code) | 57.2Reported best | 55.4≠Thinking High · tools | +1.8 pts#48 vs #52 of 85 |
| CritPtScientific Reasoning · Score | 8.0Thinking · no tools | 5.7≠Thinking High · no tools | 1.40×#62 vs #66 of 134 |
| GPQA DiamondScientific Reasoning · Accuracy | 90.5Thinking · no tools | 87.3≠Thinking · tools | +3.2 pts#46 vs #69 of 200 |
| MedCodeClinical Workflows · Accuracy (%) | 40.1Unspecified Reasoning · tools | 49.6≠Thinking High · tools | −9.5 pts#63 vs #21 of 97 |
| MedScribeClinical Workflows · Rubric pass rate (%) | 78.2Unspecified Reasoning · tools | 83.7≠Thinking High · tools | −5.5 pts#61 vs #42 of 99 |
| BrowseCompFact Finding · Accuracy | 83.2Thinking · tools + web | 54.9≠Thinking · tools | +28.3 pts#20 vs #45 of 61 |
| EBR-benchMemory & Persistence · Score (% of 21 objectives, best of the two scored playthroughs) | 2.4Thinking · tools | 12.7≠Thinking High · tools | −10.3 pts#29 vs #23 of 30 |
| MMMU-ProVisual Understanding · Accuracy | 79.4Thinking · no tools | 78.4Thinking · no tools | +1.0 pts#36 vs #45 of 136 |
| Creative WritingWriting | 1721Normal · no tools | 1626Normal · no tools | 1.06×#30 vs #42 of 121 |
| SAGEService Workflows · Accuracy (%) | 50.2Unspecified Reasoning · tools | 43.7≠Thinking High · tools | +6.5 pts#17 vs #47 of 82 |
| τ³-BankingService Workflows · Score | 23.3Thinking · tools | 22.1≠Thinking High · tools | 1.05×#51 vs #55 of 120 |
| HLEKnowledge Exams · Accuracy | 54.0Thinking · tools + web | 35.2≠Thinking · tools | +18.8 pts#33 vs #82 of 151 |
| ECICapability Indices · ECI score (capability index, higher is better) | 151Reported best | 150Reported best | 1.01×#45 vs #49 of 168 |
Repository Engineering · Accuracy
Repository Engineering · Accuracy
Agentic Development · Accuracy
Code Generation & Editing · Pass rate (%)
Mathematics · Accuracy (verification_code)
Mathematics · Accuracy (verification_code)
Scientific Reasoning · Score
Scientific Reasoning · Accuracy
Clinical Workflows · Accuracy (%)
Clinical Workflows · Rubric pass rate (%)
Fact Finding · Accuracy
Memory & Persistence · Score (% of 21 objectives, best of the two scored playthroughs)
Visual Understanding · Accuracy
Writing
Service Workflows · Accuracy (%)
Service Workflows · Score
A dash means no result is available.
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
Kimi K2.6 costs55%lessthan GPT-5 per million tokens — 2.2× cheaper.Using international rates vs standard rates at this token mix.
Kimi K2.6: International pricing. Kimi Intl standard
GPT-5: Standard pricing. OpenAI official pricing (developers.openai.com/api/docs/pricing) verified 2026-11-13; standard tier; Deprecated; shutdown 2026-12-11 (replacement: gpt-6.1-sol)
| Price / 1M tokens | Kimi K2.6 | GPT-5 | Ratio |
|---|---|---|---|
| Input | $0.95 | $1.25 | 0.76× |
| Output | $4.00 | $10.00 | 0.40× |
| Cached input | $0.16 | $0.125 | 1.28× |
| Cache write | — | — | — |
| Blended | $0.991 | $2.19 | 0.45× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
| Benchmark | Kimi K2.6 | GPT-5 |
|---|---|---|
| Agents' Last Exam v1.0 | Not reported | |
| AIME 2026 | Not reported | |
| ARC-AGI-1 | Not reported | 1 more settings |
| ARC-AGI-2 | Not reported | 1 more settings$0.4486 / task Average · ARC-AGI-2 Score: 7.5 Thinking Medium (No Tools) |
| AutomationBench | $0.28 / task Average · Pass Rate: 2.6 Reported best (effort unspecified) · zapier-private-heldout | Not reported |
| Code Migration | $5.1121 / task Average · Hidden-test pass rate (%): 27.8 Unspecified Reasoning (With Tools) | Not reported |
| EMB (Excel Modeling) | Not reported | |
| Finance Agent v2 | Not reported | |
| Legal Research Bench | Not reported | |
| MedCode | Not reported | |
| MedScribe | Not reported | |
| Public Benefits Bench v1.1 | $0.9063 / task Average · Criterion pass rate (%): 56.6 Unspecified Reasoning (With Tools) | Not reported |
| SAGE | Not reported | |
| Tax Agent Bench | $0.5655 / task Average · Accuracy (%, rubric × citation quality): 12.4 Unspecified Reasoning (With Tools) | Not reported |
| Vibe Code Bench v1.1 | Not reported |
| Attribute | Kimi K2.6 | GPT-5 |
|---|---|---|
| Availability | Available | Deprecated |
| Provider | Moonshot AI | OpenAI |
| Model type | Reasoning model | Foundation model |
| First release | April 20, 2026 | August 7, 2025 |
| Context window | 256K tokens | 400K tokens |
| Max output | — | 128K tokens |
| Total parameters | 1.0T | — |
| Active parameters | 32B | — |
| Architecture | Mixture of Experts (MoE) | Undisclosed |
| Documented capabilities | Reasoning, Multilingual, Coding, Tool use, Function calling, Computer use, Long context | Reasoning, Multilingual |
| Knowledge cutoff | 2025-04 | — |
| Input modalities | Text, Image, Video | Text, Image |
| Output modalities | Text | Text |
| Weights | Open weights | Closed / API only |
| Licence | Modified MIT | — |
| Commercial use | Commercial use allowed | Proprietary |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Kimi K2.6 or GPT-5? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.